Unsupervised Learning of Finite Mixture Models
IEEE Transactions on Pattern Analysis and Machine Intelligence
Extensions to the k-Means Algorithm for Clustering Large Data Sets with Categorical Values
Data Mining and Knowledge Discovery
Path-Based Clustering for Grouping of Smooth Curves and Texture Segmentation
IEEE Transactions on Pattern Analysis and Machine Intelligence
Cluster ensembles: a knowledge reuse framework for combining partitionings
Eighteenth national conference on Artificial intelligence
Combining Multiple Weak Clusterings
ICDM '03 Proceedings of the Third IEEE International Conference on Data Mining
Combining Multiple Clusterings Using Evidence Accumulation
IEEE Transactions on Pattern Analysis and Machine Intelligence
Clustering Ensembles: Models of Consensus and Weak Partitions
IEEE Transactions on Pattern Analysis and Machine Intelligence
Toward Objective Evaluation of Image Segmentation Algorithms
IEEE Transactions on Pattern Analysis and Machine Intelligence
Constrained spectral clustering via exhaustive and efficient constraint propagation
ECCV'10 Proceedings of the 11th European conference on Computer vision: Part VI
A theoretic framework of K-means-based consensus clustering
IJCAI'13 Proceedings of the Twenty-Third international joint conference on Artificial Intelligence
| Hi-index | 0.01 |
This paper presents a fast simulated annealing framework for combining multiple clusterings (i.e. clustering ensemble) based on some measures of agreement between partitions, which are originally used to compare two clusterings (the obtained clustering vs. a ground truth clustering) for the evaluation of a clustering algorithm. Though we can follow a greedy strategy to optimize these measures as objective functions of clustering ensemble, some local optima may be obtained and simultaneously the computational cost is too large. To avoid the local optima, we then consider a simulated annealing optimization scheme that operates through single label changes. Moreover, for these measures between partitions based on the relationship (joined or separated) of pairs of objects such as Rand index, we can update them incrementally for each label change, which makes sure the simulated annealing optimization scheme is computationally feasible. The simulation and real-life experiments then demonstrate that the proposed framework can achieve superior results.